Discussion
In this study, we aimed to explore whether baseline vagally-mediated HRV (measured via RMSSD) is a marker of the habitual tendency to ruminate, which is a maladaptive thought process that makes individuals vulnerable to mental disorders (measured by the RRS), in a large sample of healthy individuals. Based on our data from over 1100 volunteers, no evidence is found for the relationship between baseline vagally-mediated HRV at rest and rumination.
Reduced resting HRV has been reported in various forms of psychopathology (Chalmers et al., 2014; Faurholt-Jepsen et al., 2017; Tak et al., 2009), and the association between reduced HRV with trait rumination has been reported in clinical and experimental studies (Johnson et al., 2014; McLaughlin & Nolen-Hoeksema, 2011). For example, a correlational study consisting of 117 women (55 recovered from depression and 56 healthy controls) found a significant relation between brooding, a maladaptive form of rumination, and HRV, such that women reporting higher levels of brooding exhibited lower levels of HRV (Woody et al., 2014). Another study monitored HRV in 52 healthy volunteers throughout the day and found that the tendency to worry (another perseverative cognition, highly correlated to rumination) was associated with lower HRV (Brosschot et al., 2007; Fresco et al., 2002). In line with these results, the seminal neurovisceral integration model proposes that in healthy individuals, the prefrontal cortex is in control over subcortical areas, resulting in tonic inhibitory control of the parasympathetic branch of the ANS (i.e., higher HRV; Thayer & Lane, 2009). As a result, these neural and autonomic regulations result in less viscous cycles of ruminative thoughts and turn higher resting HRV (Song et al., 2022). However, in the current study, this association between the tendency to ruminate and baseline vagally-mediated HRV measured at rest cannot be observed in healthy individuals, even after controlling for gender, age, and the level of depressive symptoms. Moreover, Bayesian analyses showed a higher likelihood of no association between HRV. Finally, our results exclude a non-linear relationship between HRV with rumination and the level of depressive symptoms, which is in line with a prior study that cannot find a linear or non-linear effect of baseline resting HRV on positive emotion and depression (Silvia et al., 2014).
A meta-analysis reviewing the relationship between physiological concomitants with perseverative cognition found that maladaptive thought processes (including rumination and worry) were associated with decreased HRV in 8 correlational studies, but the literature was insufficient to conclude whether this association existed specifically for both traits (i.e., habitual tendency) and state (i.e., momentary) rumination (Ottaviani et al., 2016). Two studies focused on the relationship between the tendency to ruminate and HRV recovery in children aged 7-14 and reported inconsistent results (Sample size: 100 and 103; Borelli et al., 2014; Gentzler et al., 2013). One study found a negative correlation between resting HRV and emotion regulation, as measured by the difficulties in emotion regulation scale, in 198 undergraduate students (Rankin et al., 2014). Only two studies described in the previous paragraph specifically focused on trait rumination, worry, and resting HRV and reported significant negative associations (Brosschot et al., 2007; Woody et al., 2014). The remaining three studies (of the 8 studies included in the meta-analysis) were all related to instantaneous changes in HRV or state rumination and worry showing significant results but in the context of laboratory-based operation cannot generalize to the effects of chronic stress in everyday life (Cash et al., 2013; Gazelle & Druhen, 2009; Mezulis et al., 2014). Thus, existing studies either have focused on all-female samples and include depression risk rather than completely healthy controls, conclusions were drawn from young school children, or the sample size was relatively low. In addition, although an overlap, there is still distinguishing characteristic of rumination and worry (Nolen-Hoeksema et al., 2008). Therefore, the finding of this study, based on a large sample size and a young healthy population with a wider age range, suggests that the tendency to ruminate is not associated with lower HRV in healthy individuals.
Some characteristics of our study can, at least in part, explain the null association between HRV and rumination. Although the participants in the current dataset report a wide range of habitual tendencies towards rumination, they are healthy young individuals without mental diseases. Thus, it is possible that the association between resting HRV and rumination is more robust in individuals who have impaired parasympathetic control due to long-term stress exposure or who suffer from stress-related mental disorders. As a systematic review suggests, individuals who reported childhood adversity (retrospect of the past, a ruminative thought) have no significant overall association with resting HRV, whereas the clinical sample, as significant moderators, may have small significant association under specific circumstances (Wesarg et al., 2022). These might provide an explanation for the absent association between trait rumination and resting HRV in the current large sample of healthy individuals. Also, HRV was measured during a resting period, and no stress or ruminative processes were induced. Thus, another explanation could be that the negative correlation is more related to state (i.e., at that moment) rumination and acute changes in HRV under experimental induction paradigms rather than in a resting phase (Ottaviani et al., 2016). In sum, based on our findings, resting vagally-mediated HRV cannot be regarded as an index of the tendency to ruminate, and thus a vulnerability factor in trait rumination, in healthy individuals.
Overall, despite an absent correlation between baseline HRV measured during rest, associations between other variables are in line with the literature. The results of our large-scale cross-sectional dataset show that the resting HRV of males was slightly higher than that of females (Hamidovic et al., 2020). Furthermore, sex differences were found in the rumination scores (females scoring higher as compared to males), but not in the questionnaires regarding depressive symptoms. In addition, our results also show that, based on a limited age range, there was a significant negative correlation between age and rumination scores, as well as age and BDI-II scores, indicating that with the increase in age, people experience fewer symptoms of rumination and depression. This observation is in line with prior studies reporting that older people show a decrease in processing negative stimuli and thus show a positive effect on emotion (Nashiro et al., 2012). Rumination and depression were also highly correlated, as previous studies suggested that rumination is a potential risk factor for depression and can prolong and intensify depressive symptoms (Nolen-Hoeksema et al., 1997; Spasojević & Alloy, 2001). Overall, these findings indicate that the association between rumination and related psychological constructs is consistent with those reported in the previous literature, showing the validity of our data.
Accurate baseline measurements are critical to measuring the impact of cognitive tasks or group designations. At the resting baseline phase, participants usually sit with their knees at a 90° angle, both feet flat on the floor, hands on thighs, and no psychological tasks. This protocol involves inevitable mind wandering, disruptive thoughts, or difficulty sitting still for some participants. A popular alternative to this forced relaxation could be vanilla baseline, where participants must perform a task that requires sustained attention but a minimal cognitive load (such as reading a magazine or coloring). Importantly, our results show that different baseline types influenced resting HRV, denoting that the HRV values during a vanilla baseline were higher as compared to a baseline in rest. Further studies should consider baseline type depending on specific experimental tasks and purpose, while there is no simple answer to defining an appropriate baseline measurement.
Some limitations of our study should be considered. It is important to mention that the sex ratio in this study is not balanced and that sex itself affects HRV, which may impede the broad applicability of our results. Body mass index is also associated with HRV, but it was not measured in some of our experiments. In addition, the measurement of trait rumination and resting HRV was inconsistent in time and space. The trait rumination questionnaire was generally completed online (before the experiment), while HRV was measured in a laboratory environment, and generally not on the same day. Moreover, as all the participants were recruited in the same lab, we cannot rule out the possibility of subjects that participated in multiple experiments. All the above factors might contribute to the observed results.
To conclude, this is the first study using a large cross-sectional dataset to examine the relationship between baseline resting HRV and trait rumination in healthy individuals. The results indicate that there is no association between HRV under a specific condition, namely baseline resting, and the tendency to ruminate in daily life, based on a sample of more than thousands of adults. Thus, vagally-mediated HRV at rest could not be regarded as a marker of trait rumination in healthy individuals.